One of the characteristics of infrared image is that it has a high dynamic range (such as the temperature difference between the ground and the sky is large), while the temperature difference between some targets and their background or local targets is relatively small. Therefore, in order to quantify the infrared scene with large dynamic range with high enough accuracy, high-performance thermal imaging systems usually use 14 bits or higher accuracy ad to sample and quantify the detector output signal. The usual display device only requires 8 bits of data width, so the high-precision data of 14 bits needs to be compressed. Therefore, if the compression method is not handled properly, the information that originally detected the large dynamic image cannot be displayed, that is, the large dynamic image compression may cause the loss of the original information, which is manifested as the loss of image details in the displayed image.

For the commonly used linear mapping (such as AGC, automatic gain control) or nonlinear mapping (such as histogram equalization, gamma transform) compression methods, because the local image of small targets or objects is not dominant in temperature difference or the number of pixels, these methods generally have defects in the compression of large dynamic images.

In recent years, with the in-depth research on this issue at home and abroad, some new ideas and algorithms have been proposed. Among them, DDE (Digital detail enhancement) technology provides a new technical way to solve this problem. DDE is an advanced nonlinear image processing algorithm. When processing the algorithm, first use a special filter to separate the background layer (low-frequency part) and detail layer (high-frequency part) of the image, and then carry out corresponding gray enhancement and noise suppression processing on the extracted background layer and detail layer respectively, Then adjust and compress the dynamic range of each layer, and finally synthesize an output image of 8 bits. In this way, the large dynamic temperature difference in the 14 bits image and the local details of the target can be displayed within the range of 8 bits.

As shown in the above figure, compared with ordinary linear mapping and histogram equalization algorithms, DDE can greatly enhance the details of the image while retaining the dynamic range. In figure (a), the local temperature difference of high-temperature objects such as human face (position 1) and circuit board (position 3) is not only reflected, but also the mirror image formed by the reflection of the thermal image's own radiation on the glass can be found in position 2. In figure (c), the furnace wall of the blast furnace and the fuel in the middle are also clearly visible after being enhanced by DDE.

The essence of DDE is to display the low contrast target in the large dynamic range infrared image. The specific implementation methods are different. The effectiveness and real-time performance of the algorithm are the key to the practicality of DDE technology. At present, huajingkang optoelectronic technology has implemented an autonomous DDE algorithm on the software and hardware platform, which can enhance the details of the infrared image and suppress the image noise at the same time, The client software and SDK secondary development kit provided by our company support DDE algorithm, which can greatly improve the display effect of infrared images in practical applications.